Search Results (6,449)

Camera calibration is crucial to accurate vision tasks for its establishment of the mapping between 3D space and 2D image space. Traditional calibration methods often suffer from limited accuracy and time-consuming processes. In this study, we propose an automatic guidance system leveraging an improved particle swarm optimization algorithm to achieve fast and high-precision camera calibration. Our system dynamically recommends optimal camera poses for the next calibration image, effectively reducing calibration uncertainty and enhancing accuracy. Furthermore, for wide-angle cameras, we introduce a pre-estimation of distortion coefficients to guide the calibration process, significantly improving the calibration of distortion parameters. Experimental results demonstrate that our method outperforms existing guidance systems, achieving higher calibration accuracy with fewer images and shorter calculation time. The results of camera parameters calibrated by the system are applied to the reconstruction based on point clouds, and can achieve desirable reconstruction effect. The proposed system holds promise for applications in film and television shooting, promoting the development of the industry by reducing calibration errors and equipment debugging time. Full article

Distribution system restoration (DSR) has become increasingly challenging due to frequent topology changes and complex nodal interactions, especially in systems with a high penetration of distributed energy resources (DERs). Existing deep reinforcement learning (DRL) methods remain limited in representing inter-nodal relationships and incorporating domain prior knowledge. Therefore, this paper proposes a graph attention-based coupling-aware reinforcement learning method. From a non-Euclidean spatial perspective, the proposed method uses the distribution power transfer factor (DPTF) to quantify the strength of electrical coupling between nodes. The resulting coupling strengths are embedded as entries of the graph adjacency matrix, allowing the model to capture complex nodal interactions driven by power transfer. An aware graph attention network (AGAT) is further developed, where adjacency matrix with prior knowledge is introduced as a bias term in the attention coefficient calculation. This design guides GAT to generate differentiated node representations enriched with physical information. Based on the extracted graph features, proximal policy optimization (PPO) is employed to determine restoration decisions. Case studies on the IEEE 34-bus system demonstrate that the proposed method outperforms benchmark algorithms in training convergence, restored power, and online decision efficiency, enabling fast and effective distribution system restoration. Full article

Aiming at the problems of traditional Rapidly exploring Random Tree Star (RRT*), its modified algorithms in the three-dimensional complex underwater path planning of autonomous underwater vehicles (AUVs), including sampling redundancy, simplistic expansion mechanism, mismatch between planned paths and AUV motion constraints, and insufficient real-time performance, this paper proposes an improved bidirectional artificial potential field RRT* (Improved BI-APF-RRT*) algorithm. The algorithm adopts a hybrid sampling strategy to concentrate sampling in high-value regions and optimize node distribution. A three-level progressive expansion strategy is designed to balance fast convergence and excellent obstacle avoidance capability. Meanwhile, a dynamic target switching strategy is introduced to enhance the coordination efficiency of bidirectional search trees. Simulation results show that in three-dimensional underwater obstacle environments with different complexity levels, the proposed algorithm outperforms comparative algorithms such as GB-RRT*, BI-RRT*, and APF-RRT* in terms of path length, planning time, number of generated nodes, and iteration times. The proposed algorithm provides an efficient and feasible technical scheme for the deep-sea autonomous navigation of AUVs, and is of great significance for promoting the engineering application of underwater unmanned systems. Full article

In aerospace applications, structural vibration can cause fatigue accumulation and shorten the service life of aircraft. This makes vibration suppression based on Macro Fiber Composite (MFC) piezoelectric composites an important research topic. Considering the asymmetric high-voltage operating range of the M-8557-P1 MFC from −500 V to 1500 V and its capacitive impedance characteristics within the 1000 Hz operating frequency band, this paper designs a laboratory prototype of a high-voltage driver. The prototype adopts a voltage–current dual closed-loop structure and a current-tracking PWM control strategy. Under the tested laboratory conditions, the prototype exhibited a relatively fast transient response and a certain dynamic driving capability for capacitive loads. Based on the laboratory prototype, an auxiliary signal-conditioning module and a digital control module equipped with an active control algorithm were further developed. These modules were integrated with the laboratory prototype to form a high-voltage closed-loop control system for MFC piezoelectric sensing and actuation. Ground laboratory tests were conducted on a high-aspect-ratio unmanned aerial vehicle wing. The experimental results show that, when the dominant vibration frequency is approximately 3.6 Hz, the response converges to a steady state within 4.77 s after control is applied. In the steady state, the root-mean-square displacement decreases from 15.57 mm to 4.28 mm, corresponding to a reduction of 72.52%. This result demonstrates the effectiveness of the active vibration control system under this representative application scenario. Full article

To address the difficulty in accurately characterizing the fault current response of renewable energy grid-connected devices during high-/low-voltage ride-through, this paper proposes a fault current response modeling and parameter identification method based on adaptive nonlinear compensation. First, with the fault voltage and pre-fault operating point as input variables, a basic quadratic equivalent model is established to describe the main variation characteristics of active and reactive currents during high-/low-voltage ride-through. Second, nonlinear compensation terms are introduced into the basic model to correct the response deviation caused by the simplification of fast electromagnetic control links in the electromechanical transient equivalent process, thereby improving the representation capability of the model for complex fault current characteristics. Furthermore, considering that the structural parameters of the nonlinear compensation terms are difficult to directly identify using the traditional least squares method, a differential evolution–ridge regression (DE–Ridge) hierarchical identification method is proposed. In this method, the differential evolution algorithm is used in the outer layer to adaptively optimize the nonlinear structural parameters, while ridge regression is used in the inner layer to solve the corresponding linear coefficients. Case study results show that, compared with the traditional quadratic equivalent model and the fixed nonlinear compensation model, the proposed method further reduces the fault current identification error on the validation set and improves the identification accuracy and generalization capability of fault current responses during high-/low-voltage ride-through. Full article

Three-dimensional (3D) path planning for Unmanned Aerial Vehicles (UAVs) presents a challenging multi-objective optimization problem that necessitates a balanced trade-off among path length, flight safety, and trajectory smoothness, especially in complex environments such as mountainous or hilly terrains. Traditional and even many meta-heuristic planning algorithms often suffer from premature convergence and suboptimal solution quality when navigating such intricate 3D spaces. To address these limitations, this paper proposes an Improved Dhole Optimization Algorithm (IDOA) that exhibits fast convergence and strong global optimization capabilities. The IDOA enhances the original DOA framework by integrating a logistic-map-based chaotic mapping, a dynamic chaotic perturbation mechanism, and an adaptive stage-division strategy. The algorithm is designed to address the 3D path planning problem for quadrotor UAVs, supporting typical flight maneuvers including climb/descent, obstacle avoidance, and smooth turning in simulated complex hilly terrain. A multi-objective fitness function incorporating path length, safety, and smoothness is designed, which constrains the optimization to generate collision-free, smooth paths that satisfy the quadrotor UAV’s dynamic maneuver constraints. Convergence curves confirm that IDOA significantly outperforms the original DOA in terms of convergence speed and final path optimality. Detailed experimental results show that compared to the original DOA, IDOA achieves a 6.31% improvement in minimum fitness values, a 9.7% reduction in average path length, and a 45.28% reduction in average path curvature when compared to the baseline DOA. These consistent performance improvements demonstrate that IDOA provides an effective and robust solution for offline pre-flight 3D path planning in complex terrain, offering valuable technical support for autonomous UAV navigation in practical application scenarios such as terrain surveying and disaster search and rescue. Full article

Chronic low-grade inflammation accompanies cardiometabolic disease, while routine risk assessment is based mainly on fasting measurements. This structured narrative review summarises human evidence for discrete postprandial changes in triglyceride-rich lipoproteins and remnants, glucose and insulin, endotoxin-related markers, innate immune cells and endothelial function. The evidence is comparatively consistent for postprandial lipaemia, glycaemic excursions and acute flow-mediated dilation responses, whereas endotoxin, cytokine and cellular findings are smaller, assay-sensitive and less consistently replicated. Based on this evidence, we introduce PRISM-CM (Postprandial Inflammatory Stress Modules in CardioMetabolic disease) as an author-developed, hypothesis-generating evidence map. It comprises five candidate biological domains—lipid–remnant burden, glucose–insulin stress, endotoxin handling, innate immune-cell activation and endothelial response—with recovery kinetics treated as a cross-domain analytic dimension. The framework was not derived by clustering, consensus methods or predictive modelling; its illustrative response patterns are not validated endotypes. A composite score is not proposed because the independence, reproducibility and incremental predictive value of the candidate measurements have not been established. Potential future diagnostic and prognostic applications require prospective validation. Reference ranges, age- and sex-specific norms, within-person reproducibility, reproducible response patterns and outcome-linked thresholds remain unknown. PRISM-CM is therefore not a clinical algorithm, risk score or treatment-selection tool. Full article

Sparse adaptive equalization is widely used for time-varying multipath channels due to its ability to exploit channel sparsity and reduce computational complexity. However, conventional normalized least mean square (NLMS)-based algorithms usually rely on fixed step-size strategies or manually selected regularization parameters, which limits their capability to simultaneously achieve fast convergence, low steady-state error, and robust tracking performance under varying channel conditions. This paper proposes a sparsity-aware variable step-size NLMS (SA-VSS-NLMS) equalizer with dynamic sparsity-adaptive regularization for time-varying sparse channels. The proposed framework employs online Hoyer sparsity estimation as a unified feedback signal to jointly adjust the adaptive step size and regularization strength. By incorporating sparsity information into both adaptation processes, the proposed method enables rapid convergence during channel variations while maintaining accurate steady-state tracking in sparse environments. In addition, an anomaly-aware tracking mechanism is introduced to improve recovery performance under abrupt channel changes. The convergence behavior and stability properties of the proposed algorithm are analyzed, and extensive simulations are conducted under stationary, time-varying, and fractional-delay sparse channel conditions. The results demonstrate that the proposed SA-VSS-NLMS achieves faster convergence, lower steady-state error, improved tracking robustness, and enhanced bit error rate (BER) performance compared with conventional NLMS-based and sparse adaptive filtering methods. Moreover, the proposed algorithm maintains linear computational complexity with respect to the filter length, making it suitable for practical adaptive communication systems. Full article

The spatial distribution of occupants shapes energy use and work-related performance. Algorithmic seating optimization can shorten communication distances, reduce electricity for lighting, and increase daily light exposure, whereas prior studies often targeted a single objective. Given the nondeterministic polynomial-time (NP)-hardness of key subproblems and the complexity of multi-objective search, this study evaluates heuristic and metaheuristic methods driven by sensor data from an open-plan office. Evolutionary, sampling-based, and model-based approaches are compared in terms of Pareto dominance, solution diversity, stability, convergence, and computation time. Results show that multi-criteria optimization with real-world data yields clear differences in performance profiles and search space exploration. Non-dominated Sorting Genetic Algorithm III (NSGA-III) contributes the highest share of Pareto solutions, while Hybrid Metaheuristics (HMH) achieves the largest target space coverage (hypervolume). Markov Chain Monte Carlo (MCMC) and Pareto Simulated Annealing (PSA) deliver particularly stable gains in light dose, while Bayesian optimization contributes no Pareto solutions in the present setting. Certain user pairings and spatial allocation patterns remain consistent across strategies, indicating persistent structural properties of the search space. Fast methods such as Deep Optimization (DO), Multi-Objective Pareto Simulated Annealing (PSA-Multi), Mulit-Objective Evolutionary Algorithm based on Decomposition (MOEA/D), and MCMC are efficient, whereas NSGA-III offers the highest solution quality at greater computational cost. These findings advance understanding and support algorithm selection and space analysis for similar combinatorial allocation problems. Full article

Magnetic leakage detection is one of the most commonly used methods for pipeline inspection, which mainly uses magnetic sensors to detect the magnetic leakage field on the internal and external surfaces of the pipeline to determine whether there are defects in the pipeline. The defect quantification algorithm includes a forward model and an optimization algorithm, in which the estimation of target defects using optimization algorithms is one of the key aspects of defect inversion. Most of the existing optimization algorithms are based on particle swarm algorithms (PSOs) and genetic algorithms (GAs), which are prone to premature problems and have low convergence accuracy. To address the problems in the process of defect inversion, this paper proposes a new inversion algorithm, which obtains part of the prior knowledge from the application context of defect inversion, and adopts the decay oscillation function as the nonlinear convergence factor based on the whale optimization algorithm (WOA). In addition, referring to the concepts of “genetic” and “mutation” in the GA, a diversity variation strategy based on dynamic step size is designed. The algorithm designed has the advantages of fast operation and high search accuracy. At the end of the paper, two sets of experiments are designed to compare the improved WOA with other existing optimization algorithms. The results demonstrate that the algorithm is significantly superior to other algorithms, both in the ideal case of simulation experiments and in the practical application of defect inversion. Full article

Fractional Brownian motion (fBm) provides a powerful stochastic framework for modeling long-range temporal dependence that cannot be represented by classical Brownian motion. This study presents a numerical and theoretical investigation of constrained fractional Brownian motion with applications to stochastic financial systems. An efficient simulation framework combining Fast Fourier Transform (FFT)-based circulant embedding and Markov Chain Monte Carlo (MCMC) sampling is developed to generate long correlated trajectories under absorbing boundary conditions. The proposed algorithm enables simulations with trajectory lengths up to L = 10⁷ while reducing the computational complexity from O (L³) for direct covariance decomposition to approximately O(L log L). Numerical results accurately reproduce the theoretical autocorrelation function of fBm and confirm the expected persistence behavior governed by the Hurst exponent. Super-diffusive regimes (H > 0.5) exhibit persistent long-range correlations and enhanced survival probabilities, whereas sub-diffusive regimes (H < 0.5) display anti-persistent dynamics and increased boundary absorption. The fractional stochastic volatility formulation captures important characteristics associated with long-memory financial systems, including persistent volatility dynamics and implied-volatility structures. The proposed biophysical-based FFT–MCMC methodology provides an accurate, scalable, and computationally efficient framework for studying constrained fractional stochastic processes and offers a foundation for future investigations of fractional volatility models and related financial applications. A conceptual Adaptive Hurst Momentum framework is briefly discussed as a possible direction for future research. Full article

The global rise in obesity and associated metabolic disorders has fueled the demand for safe, multi-target therapeutics derived from natural sources. In this study, an artificial neural network combined with a genetic algorithm (ANN-GA), was employed to optimize the ultrasound-assisted extraction of loquat leaf extract (LLE). The optimal parameters—0.6% Tween 80, 328 W ultrasonic power, and 49 °C—enhanced extraction efficiency while minimizing energy consumption and surfactant use. In vitro, LLE exhibited potent dose-dependent inhibition of α-glucosidase (IC₅₀ = 3.75 μg/mL, 21.6-fold stronger than acarbose) and pancreatic lipase (IC₅₀ = 78.3 μg/mL), indicating strong inhibitory activity against these digestive enzymes. In high-fat diet-induced obese mice, LLE supplementation dose-dependently reduced body weight gain, serum total cholesterol, triglycerides, low-density lipoprotein cholesterol (LDL-C), and fasting blood glucose while elevating high-density lipoprotein cholesterol (HDL-C). In addition, LLE ameliorated liver injury (decreased ALT/AST) and oxidative stress (lowered MDA, increased SOD). These findings highlight LLE as a promising multi-target natural product for managing obesity and related metabolic disorders. Full article

In order to improve the bird strike avoidance management at airports, and to realize the linkage of bird detection radar and a variety of bird repellent equipment, intelligent bird repellent decision methods have attracted widespread attention. In this paper, we are committed to exploring bird target recognition approaches, so as to prepare for the subsequent bird repellent decision methods. Kernel methods are well known to be effective in dealing with nonlinear machine learning problems. However, the performance of kernel methods heavily relies on the choice of kernel parameters. To mitigate this problem, we add the ideas of multiple kernels and randomized techniques to kernel discriminant analysis (KDA), and propose a randomized and fast multiple kernel discriminant analysis (RM-KDA) method with applications to bird recognition. The main contributions of our work are two-fold. First, we propose an innovative multiple kernel discriminant analysis (M-KDA) framework to effectively solve recognition tasks with nonlinear data structures. Second, we develop a fast and efficient randomized algorithm to solve the M-KDA model without explicitly forming and storing the base kernel matrices and the ensemble kernel matrix in advance, which greatly reduces the computation and storage requirements. The extensive experiments on benchmark recognition databases demonstrate the effectiveness of the proposed algorithm. Full article

The rapid shutdown of high-head and large-capacity Pelton turbines equipped with long penstocks often induces severe water hammer effects, threatening the structural safety of the diversion system. Traditional linear or simple piecewise nozzle closing laws struggle to achieve an optimal balance between shortening the nozzle closing time and suppressing the maximum transient pressure. To address this issue, this study proposes a multi-objective optimization framework for the non-linear nozzle-closing strategy of a large-capacity six-nozzle Pelton turbine. A one-dimensional transient flow model of the complex diversion system was established and solved using the Method of Characteristics (MOC). Subsequently, the Non-dominated Sorting Genetic Algorithm II (NSGA-II) was coupled with the model to globally optimize the discretized 10-segment nozzle-closing trajectory. The optimization objectives were to minimize both the total closing time and the maximum pressure before the nozzle. The results demonstrate that the NSGA-II algorithm autonomously converges to a “Fast-Slow-Fast” non-linear closing pattern. Compared to the baseline linear closing strategy, the optimized trajectory reduces the total shutdown time by 7.22% (from 18.15 s to 16.84 s) and decreases the maximum pressure before the nozzle by 7.61 m, effectively mitigating the net water hammer pressure rise by 13.40%. Physical mechanism analysis reveals that the intermediate “slow” phase effectively staggers the constructive superposition of reflected positive water hammer waves, achieving an active peak-shifting effect. This study can provide a theoretical reference for the safe and efficient shutdown control of large-capacity Pelton turbines. Full article

To ensure the guidance accuracy and flight safety of instrument landing systems (ILSs), it is imperative to mitigate multipath interference caused by reflections from airport structures, whose core detrimental effect is the excess deviation of the difference in depth of modulation (DDM). This paper presents an automated design methodology with the explicit objective of directly minimizing DDM, employing a genetic algorithm (GA) to optimize additional metallic baffles adjacent to a building, thereby achieving a “stealth” effect for the building structure. The method encodes the layout parameters of additional metallic baffles adjacent to a building into chromosomes, searching for the optimal configuration through iterative evolution. Each generation applies the efficient half-space multilevel fast multipole method (HSMLFMM)—for the first time in ILS interference simulation—to accurately compute the radiation field of every candidate design. The maximum resultant DDM along the glide path serves as the fitness function for selection. Optimization and validation are conducted for four typical scenarios where the interference source is located 50, 100, 150, and 200 m from the runway centerline. The optimized DDM values are reduced to 4.49, 4.78, 4.63, and 4.87, respectively, all below the ICAO Annex 10 tolerance limit of 5$\mathsf{\mu }$A for CAT III operations. The corresponding reduction percentages are 75.85%, 83.93%, 90.73%, and 90.19%, with a maximum reduction of 90.73% achieved in the 150 m scenario. This research establishes an efficient, automated closed-loop optimization workflow, which offers a viable approach for the intelligent and precision design of low-observable buildings at airports. Full article